As I understood that article, it seemed to be saying that formal methods often don't provide a good enough cost/benefit tradeoff to be worth it in many situations to many organizations. Right at the end, it basically says that your favorite underappreciated formal method isn't used because it costs too much and/or delivers too little. If an outfit that sells formal methods is saying that, then I suspect that (many) formal methods actually do cost too much for what benefit they deliver.
A key point they made is that formal methods, as currently practiced, have a cost-benefit curve that is very hockey-shaped.
This doesn't really mean that the cost isn't worth it, but rather that it's hard to demonstrate the cost-effectiveness in a small pilot.
It also means that the pay-off is fragile. Because apparently a lot of things need to come together for the value to show.
